Introduction to Econometrics

Professor Yongmiao Hong

September 22, 2024
Copyright ◎ 2024 by Professor Hong Yongmiao, All rights reserved. Requests for permission should be mailed to: [email protected]
1. 版权归作者洪永淼教授所有；
2. 不得移除作者署名，否则将视为侵权；
3. 对于不遵守此声明或者其他违法使用本文内容者，作者依法保留追究权等。
4. 发现课件错误请联系作者 [email protected]
 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 2

Introduction

• Econometrics has become an integrated part of teaching and

research in modern economics.
• Core courses in economics:

Microeconomics Macroeconomics

Financial
Econometrics
economics

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 3

Introduction

• Empirical revolution in the past 40 years in economics:

Data-based empirical
studies in economics

Evidence-based causal
inference

Consistent with scientific

research paradigm

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 4

Introduction

• Key differences of empirical studies between economics and

natural sciences:

Controlled experiments in natural sciences

Nonexperimental nature of economic system and

economic data

• The importance of econometrics as a general methodology of

empirical research in economics has been increasingly
recognized over the past several decades.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 5

Introduction

• Big data in economics:

Traditional structured data

New structured data: (e.g.) interval-valued data,

symbolic data, functional data

Unstructured data, e.g., text data, photo data, video

data, audio data

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 6

Introduction

• New forms of data call for innovation and development of

theory, methods and tools in modern econometrics.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 7

 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 8

Quantitative Features of Modern Economics

Step 1: Data collections and summary of empirical stylized facts

Economic Empirical
Data
theories validation Applications
collections
/models /inference

Data collections:
✓ Surveys
✓ Field Studies
✓ Experimental economics
✓ Big Data

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 9

Quantitative Features of Modern Economics

Step 1: Data collections and summary of empirical stylized facts

The so-called stylized facts are often summarized from observed
economic data.
Example 1：Engel curve in microeconomics

The share of a consumer’s

expenditure on a commodity
out of her or his total income
will vary as his/her income
changes.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 10

Quantitative Features of Modern Economics

Step 1: Data collections and summary of empirical stylized facts

Example 2: Phillips Curve in macroeconomics

A negative correlation
between the inflation rate
and the unemployment rate
in an aggregate economy.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 11

Quantitative Features of Modern Economics

Step 1: Data collections and summary of empirical stylized facts

Example 3: Volatility Clustering in Finance

A high volatility today tends to be followed by another high

volatility tomorrow, a low volatility today tends to be followed by
another low volatility tomorrow, and both alternate over time.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 12

Quantitative Features of Modern Economics

Step 1: Data collections and summary of empirical stylized facts

The empirical A starting point for

serve as economic research
stylized facts

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 13

Quantitative Features of Modern Economics

Step 2: Development of economic theories/models

Economic Empirical
Data
theories validation Applications
collections
/models /inference

● With the empirical stylized facts in mind, economists then

develop an economic theory or model.
● This usually calls for specifying a mathematical model of
economic theory.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 14

Quantitative Features of Modern Economics

Step 2: Development of economic theories/models

Example 1: Classical Keynesian Theory—Keynes (1936)

The classical Keynesian theory can be summarized by two simple

mathematical equations:

National income identity: 𝑌 = 𝐶 + 𝐼 + 𝐺

Consumption function: 𝐶 = 𝛼 + 𝛽𝑌
Where:
𝑌is income, 𝐺is government spending,
𝐶is consumption, 𝛼is the “survival level”,
𝐼is private investment, 𝛽is the marginal propensity to consume.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 15

Quantitative Features of Modern Economics

Step 2: Development of economic theories/models

Example 1 (Cont.): Classical Keynesian Theory—Keynes (1936)

Multiplier of income with respect to government spending :

𝜕𝑌 1
=
𝜕𝐺 1 − 𝛽

Thus, the Keynesian theory can be effectively summarized by two

mathematical equations.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 16

Quantitative Features of Modern Economics

Step 2: Development of economic theories/models

Example 2: General Equilibrium—Walras (1874)

The general equilibrium analysis addresses whether there exists an
equilibrium price vector 𝑃∗ such that all markets are clear
simultaneously:
𝐷𝑖 𝑃∗ = 𝑆𝑖 𝑃∗ for all 𝑖 ∈ {1, … , 𝑛ሽ
where:
✓ 𝑛 is the number of goods;
✓ 𝐷𝑖 𝑃 is the demand for good 𝑖;
✓ 𝑆𝑖 𝑃 is the supply for good 𝑖;
✓ 𝑃 = 𝑃1 , 𝑃2 , … , 𝑃𝑛 is a price vector for 𝑛 goods.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 17

Quantitative Features of Modern Economics

Step 3: Empirical verification of economic models

Economic
Data Empirical
theories Applications
collections verification
/models

• A key is to transform an economic model into a testable empirical

econometric model.
• One often has to assume some functional form, up to some
unknown model parameters.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 18

Quantitative Features of Modern Economics

Step 3 (Cont.): Empirical verification of economic models

• One needs to estimate unknown model parameters and make
inferences based on observed data, and check whether the
econometric model is “adequate”.
• An adequate model should be at least consistent with the empirical
stylized facts, particularly with the empirical stylized facts.

◆ Question
◆ Questions:
◼ What is the consequence of model misspecification?
◼ What is model risk?

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 19

Quantitative Features of Modern Economics

Step 4: Applications
Economic
Data Empirical
theories Applications
collections verification
/models
After an econometric model passes the empirical evaluation, it can
be used to:

Test economic theories or hypotheses

Forecast future evolution of the economy

Conduct policy evaluation

Make policy recommendation

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 20

Quantitative Features of Modern Economics

Modern economics has two important features:

Mathematical modeling for economic theory

Statistical-based empirical analysis for economic phenomena

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 21

 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 22

Mathematical Modeling

◆ Question
Why do we need mathematics and mathematical models in
economics?
⚫ There are many ways or tools to present economic theory:

Graphical Verbal Mathematical

methods descriptions models

⚫ Mathematics is just one of them.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 23

Mathematical Modeling

⚫ It has been a long history to use mathematics in economics.

In his Mathematical Principles of the

Wealth Theory, Cournot (1838) was
among the earliest to use mathematics in
economic analysis.

Walras (1874), a mathematical

economist, heavily used mathematics to
develop his general equilibrium theory.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 24

Mathematical Modeling

⚫ It has been a long history to use mathematics in economics.

✓ The game theory, which was proposed by
Von Neumann and Morgenstern (1944)
and now becomes a core in modern
microeconomics, originated from a
branch in mathematics.
• Mathematical modeling is a necessary path to empirical
verification of an economic theory.
• One needs to formulate economic theory into an empirically
testable mathematical model whose functional form or important
structural model parameters will be estimated from the observed
data.
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 25
 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 26

Empirical Validation

◆ Question
The use of mathematics, although it can ensure logical
consistency of a theory itself, cannot ensure that economics is a
science.

 An economic theory would be useless from a practical point

of view if the underlying assumptions are incorrect or
unrealistic.
 To be a science, an economic theory must be consistent with
reality. It must be able to explain historical stylized facts and
predict future economic phenomena.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 27

Empirical Validation

◆ Question
How to check a theory or model empirically? Or how to
validate an economic theory?

 In practice, it is rather difficult or even impossible to check

whether the underlying assumptions of an economic theory or
model are correct.
 Nevertheless, one can confront the implications of an economic
theory with the observed data to check if they are consistent.
For rigorous empirical analysis, we need to use econometrics.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 28

Empirical Validation

◆ Question
How to check a theory or model empirically? Or how to
validate an economic theory?

 Econometrics has witnessed a rather rapid development in the

past several decades, due to rapid development of computing
power and availability of economic data.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 29

Empirical Validation

Fundamental Axioms of Modern Econometrics

Modern econometrics is essentially built upon the following

fundamental axioms:

Any economy can be viewed as a stochastic process governed by

some probability law.

Economic phenomenon, as often summarized in form of data,

can be reviewed as a realization of this stochastic data generating
process.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 30

Empirical Validation

Implications

One should not hope to determine precise, deterministic

economic relationships, as do the models of demand, production,
and aggregate consumption in standard micro- and macro-
economic textbooks.

No model could encompass the myriad essentially random

aspects of economic life (i.e., no precise point forecast is
possible, using a statistical terminology).

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 31

Empirical Validation

Implications

The purpose of econometrics is to infer the probability law of the

economic system using observed data.

One can test economic theory or economic hypotheses by

checking the validity of these restrictions.

There exist differences between statistical relationships and

economic causal relationships.

Econometrics is not simply an application of a general theory of

mathematical statistics to economic data.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 32

 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 33

Illustrative Examples

Econometrics can play the following roles in economics:

 Examine how well an economic theory can explain historical

economic data (particularly the important empirical stylized facts)

 Test validity of economic theories and economic hypotheses

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 34

Illustrative Examples

 Predict the future evolution of the economy

 Conduct policy evaluation and make policy recommendation.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 35

Illustrative Examples

Example 1.1: Keynesian Model, the Multiplier and Policy

Recommendation

The Simplest Keynes Model

𝑌𝑡 = 𝐶𝑡 + 𝐼𝑡 + 𝐺𝑡
ቊ
𝐶𝑡 = 𝛼 + 𝛽𝑌𝑡 + 𝜀𝑡

✓ 𝑌𝑡 is aggregate income,
✓ 𝐶𝑡 is private consumption,
✓ 𝐼𝑡 is private investment,
✓ 𝐺𝑡 is government spending,
✓ 𝜀𝑡 is consumption shock.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 36

Illustrative Examples

Example 1.1 (Cont.) : Keynesian Model, the Multiplier and

Policy Recommendation
The parameters 𝛼 and 𝛽 can have appealing economic
interpretations:
✓ 𝛼 is survival level consumption,
✓ 𝛽 is the marginal propensity to consume (MPC).

Multiplier of income with respect to government spending:

𝜕𝑌𝑡 1
=
𝜕𝐺𝑡 1 − 𝛽
which depends on the marginal propensity to consume 𝛽.
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 37
Illustrative Examples

Example 1.1 (Cont.) : Keynesian Model, Multiplier and Policy

Recommendation
⚫ Economic theory never tells exactly what 𝛽 should be for a
given economy.
✓ It is conceivable that 𝛽 differs from country to country,
because cultural factors may have impact on the
consumption behavior of an economy.
✓ It is also conceivable that 𝛽 will depend on the stage of
economic development in an economy.
⚫ Fortunately, econometrics offers a feasible way to estimate 𝛽
from observed data.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 38

Illustrative Examples

Example 1.2: Rational Expectations and Dynamic Asset Pricing

Models

Suppose a representative agent has a constant relative risk

aversion utility

𝑈 = ∑𝑛𝑡=0 𝛽𝑡 𝑢 𝐶𝑡
𝛾
𝑛 𝑡
𝐶𝑡 −1
= ∑𝑡=0 𝛽
𝛾
where
✓ 𝛽 > 0 is the agent's time discount factor,
✓ 𝛾 ≥ 0 is the risk aversion parameter,
✓ 𝑢 ∙ is the agent's utility function in each time period,
✓ 𝐶𝑡 is consumption during period t.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 39

Illustrative Examples

Example 1.2 (Cont.): Rational Expectations and Dynamic Asset

Pricing Models
• Let the information available to the agent at time 𝑡 be represented
by 𝐼𝑡 , and let 𝑅𝑡 = 𝑃𝑡 /𝑃𝑡−1 be the gross return to an asset acquired
at time 𝑡 − 1 at a price of 𝑃𝑡−1 . The agent’s optimization problem
is to choose a sequence of consumptions {𝐶𝑡 } over time to
max 𝐶𝑡 𝐸 𝑈
subject to the intertemporal budget constraint
𝐶𝑡 + 𝑃𝑡 𝑞𝑡 ≤ 𝑊𝑡 + 𝑃𝑡 𝑞𝑡−1 ,
where 𝑞𝑡 is the quantity of the asset purchased at time 𝑡 and 𝑊𝑡 is
the agent’s period 𝑡 income.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 40

Illustrative Examples

Example 1.2 (Cont.): Rational Expectations and Dynamic Asset

Pricing Models
Define the marginal rate of intertemporal substitution.
𝜕𝑢 𝐶𝑡+1 𝛾−1
𝜕𝐶𝑡+1 𝐶𝑡+1
MRS𝑡+1 (𝜃) = =
𝜕𝑢 𝐶𝑡 𝐶𝑡
𝜕𝐶𝑡

where model parameter vector 𝜃 = (𝛽, 𝛾)′

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 41

Illustrative Examples

Example 1.2 (Cont.): Rational Expectations and Dynamic Asset

Pricing Models

First order condition (FOC) of the agent's optimization

problem:
𝐸 𝛽MRS𝑡+1 (𝜃)𝑅𝑡+1 ∣ 𝐼𝑡 = 1

That is, the marginal rate of intertemporal substitution

discounts gross returns to unity. This FOC is usually called
the Euler equation of the economic system.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 42

Empirical Validation

◆ Question
How to estimate this model? How to test validity of a rational
expectations model?

 Econometricians have developed a consistent estimation method

based on the conditional moment condition or the Euler equation.
The method is named generalized method of moments (see
Hansen 1982).

 To resolve equity premium puzzle, effort has been devoted to the

development of new economic models with time-varying, large
risk aversion.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 43

Illustrative Examples

Example 1.3: The Production Function and Hypothesis on

Constant Return to Scale

Production function:
𝑌𝑖 = 𝑒𝑥𝑝 𝜀𝑖 𝐹 𝐿𝑖 , 𝐾𝑖
where
✓ 𝑌𝑖 is output of firm 𝑖,
✓ 𝐿𝑖 is labor of firm 𝑖,
✓ 𝐾𝑖 is capital of firm 𝑖,
✓ 𝜀𝑖 is a shock (e.g., uncertain weather condition if 𝑌𝑖 is an
agricultural product).

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 44

Illustrative Examples

Example 1.3 (Cont.): The Production Function and Hypothesis

on Constant Return to Scale

Constant return to scale (CRS):

𝜆𝐹 𝐿𝑖 , 𝐾𝑖 = 𝐹 𝜆𝐿𝑖 , 𝜆𝐾𝑖 for all 𝜆 > 0

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 45

Illustrative Examples

Example 1.3 (Cont.): The Production Function and Hypothesis

on Constant Return to Scale

A conventional approach to testing CRS:

• Assume the production function is a Cobb-Douglas function:
𝛽
𝑌𝑖 ≡ 𝐹 𝐿𝑖 , 𝐾𝑖 = 𝐴exp 𝜀𝑖 𝐿𝛼𝑖 𝐾𝑖
• Then CRS becomes a mathematical restriction on
parameters (𝛼, 𝛽)
𝐇0 : 𝛼 + 𝛽 = 1
• If 𝛼 + 𝛽 > 1, the production technology displays IRS.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 46

Illustrative Examples

Example 1.3 (Cont.): The Production Function and Hypothesis

on Constant Return to Scale

 In statistics, a popular procedure to test one-dimensional

parameter restriction is Student's t-test. Unfortunately, this
procedure is not suitable for many cross-sectional economic data,
which usually display conditional heteroskedasticity. One needs to
use a robust, heteroskedasticity-consistent test procedure,
originally proposed in White (1980).
 It should be emphasized that CRS is equivalent to the statistical
hypothesis 𝐇0 : 𝛼 + 𝛽 = 1 under the assumption that the
production technology is a Cobb-Douglas function.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 47

Illustrative Examples

Example 1.3 (Cont.): The Production Function and Hypothesis

on Constant Return to Scale

• This additional condition on the production function is not part of

CRS hypothesis and is called an auxiliary assumption.
• If the auxiliary assumption is incorrect, the statistical hypothesis
𝐇0 : 𝛼 + 𝛽 = 1 will not be equivalent to CRS.
• Correct model specification is essential for a valid conclusion and
interpretation for econometric inference.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 48

Illustrative Examples

Example 1.4: Effect of Economic Reforms on Transitional

Economy

Extended Cobb-Dauglas production function (after taking a

logarithmic operation)
ln 𝑌𝑖𝑡 = ln 𝐴𝑖𝑡 + 𝛼ln 𝐿𝑖𝑡 + 𝛽ln 𝐾𝑖𝑡 + 𝛾Bonus𝑖𝑡 + 𝛿Contract𝑖𝑡 + 𝜀𝑖𝑡
Where:
✓ 𝑖 is the index for firm 𝑖 ∈ {1, … , 𝑁ሽ and 𝑡 is the index for year
𝑡 ∈ {1, … , Tሽ .
✓ Bonus𝑖𝑡 is proportion of bonus out of total wage bill,
✓ Contract𝑖𝑡 is proportion of workers who have signed a fixed-
term contract.
This is an example of the so-called panel data model (see, e.g.,
Hsiao 2003).
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 49
Illustrative Examples

Example 1.4 (Cont.): Effect of Economic Reforms on Transitional

Economy
• Paying bonuses and fixed-term contracts were two innovative
incentive reforms in Chinese state-owned enterprises in 1980s,
compared to the fixed wage and life-time employment systems in the
pre-reform era (see Groves, Hong, McMillan and Naughton 1994).
• To examine effects of these incentive reforms, we consider the 𝐇0 :

𝐇0 : 𝛾 = 𝛿 = 0

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 50

Illustrative Examples

Example 1.4 (Cont.): Effect of Economic Reforms on Transitional

Economy
• It appears that the classical t-tests or F-tests would serve
our purpose here, if we can assume conditional
homoskedasticity.
• Unfortunately, this cannot be used because there may
exist the other way of causation from 𝑌𝑖𝑡 to 𝐵𝑜𝑛𝑢𝑠𝑖𝑡 : a
productive firm may pay its workers higher bonuses
regardless of their efforts. This will cause correlation
between the bonuses and the error term 𝑢𝑖𝑡 , rendering the
OLS estimator inconsistent and invalidating the classical
t-tests or F-tests.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 51

Illustrative Examples

Example 1.4 (Cont.): Effect of Economic Reforms on Transitional

Economy

• Fortunately, econometricians have developed an important

estimation procedure called Instrumental Variables (IV)
estimation, which can effectively filter out the impact of the
causation from output to bonus and obtain a consistent
estimator for the bonus parameter.
✓ IV regression has been a popular methodology to identify
economic causal relationships based on nonexperimental
observations.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 52

Illustrative Examples

Example 1.4 (Cont.): Effect of Economic Reforms on Transitional

Economy

In evaluating the effect of economic reforms, we have turned an

economic hypothesis into a statistical hypothesis
𝐇0 : 𝛿 = 𝛾 = 0

• When the hypothesis 𝐇0 : 𝛿 = 𝛾 = 0 is not rejected, we should

not conclude that the reforms have no effect. This is because the
extended production function model, where reforms are
specified additively, is only one of many ways to check effect of
reforms.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 53

Illustrative Examples

Example 1.4 (Cont.): Effect of Economic Reforms on Transitional

Economy

• For example, one could also specify the model such that the
reforms affect the marginal productivities of labor and capital
(i.e., the coefficients of labor and capital).
✓ Thus, when the hypothesis 𝐇0 : 𝛿 = 𝛾 = 0 is not rejected, we
can only say that we do not find evidence against the
economic hypothesis that the reforms have no effect. We
should not conclude that the reforms have no effect.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 54

Illustrative Examples

Example 1.5: Efficient Market Hypothesis (EMH) and

Predictability of Financial Returns

Weak form of efficient market hypothesis (EMH):

It is impossible to predict future stock returns using the history of
past stock returns:
❖ 𝐸 𝑌𝑡 ∣ 𝐼𝑡−1 = 𝐸 𝑌𝑡 ,
where
 𝑌𝑖𝑡 is the stock return in period 𝑡,
 𝐼𝑡−1 = 𝑌𝑡−1 , … , 𝑌1 is the information set at time 𝑡 − 1,
 𝐸 𝑌𝑡 |𝐼𝑡−1 : is the expected return that can be obtained when one
is fully using the information available at time 𝑡 − 1,
 𝐸 𝑌𝑡 : is the expected market average return in the long-run; it is
the expected return of a buy-and-hold trading strategy.
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 55
Illustrative Examples

Example 1.5 (Cont.): Efficient Market Hypothesis (EMH) and

Predictability of Financial Returns

• When EMH holds, the past information of stock returns has no

predictive power for future stock returns.
✓ An important implication of EMH is that mutual fund
managers will have no informational advantage over layman
investors.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 56

Illustrative Examples

Example 1.5 (Cont.): Efficient Market Hypothesis (EMH) and

Predictability of Financial Returns
◆ Question
How to test EMH?
• One simple way to test EMH is to consider autoregression:

𝑝
𝑌𝑡 = 𝛼0 + ∑𝑗=1 𝛼𝑗 𝑌𝑡−𝑗 + 𝜀𝑡
where 𝑝 is a pre-selected number of lags, 𝜀𝑡 is a random disturbance.
• EMH implies 𝐇0 : 𝛼1 = 𝛼2 = ⋯ = 𝛼𝑝 = 0.
✓ Any nonzero coefficient 𝛼𝑗 , 1 ≤ 𝑗 ≤ 𝑝, is evidence against EMH.
Thus, to test EMH, one can test whether the 𝛼𝑗 are jointly zero.
✓ The classical 𝐹-test can be used to test the hypothesis 𝐇0 when
var 𝜀𝑡 |𝐼𝑡−1 = 𝜎 2 , i.e., there exists conditional homoscedasticity.
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 57
Illustrative Examples

Example 1.5 (Cont.): Efficient Market Hypothesis (EMH) and

Predictability of Financial Returns
• However, EMH may coexist with volatility clustering ( i.e.,
var 𝜀𝑡 |𝐼𝑡−1 may be time-varying), which is one of the most
important empirical stylized facts of financial markets (see Chen
and Hong (2003) for more discussion). This implies that the
standard 𝐹-test statistic cannot be used, even asymptotically.
• Like the discussion in Subsection 5.4, when one rejects the 𝐇0
that the 𝛼𝑗 are jointly zero, we have evidence against EMH.
• However, when one fails to reject 𝐇0 that 𝛼𝑗 are jointly zero, one
can only conclude that we do not find evidence against EMH.
• The reason is that the linear 𝐴𝑅(𝑝) model is one of many
possibilities to check EMH (see, e.g., Hong and Lee 2005).

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 58

Illustrative Examples

Example 1.6: Volatility Clustering and ARCH Models

Since 1970s, oil crisis, the floating foreign exchanges system, and
the high interest rate policy in the U.S. have stimulated a lot of
uncertainty in the world economy. Economic agents have to
incorporate these uncertainty in their decision-making. How to
measure uncertainty has become an important issue.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 59

Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

In economics, volatility is a key instrument for measuring

uncertainty and risk in finance. This concept is important to
investigate information flows and volatility spillover, financial
contagions between financial markets, options pricing, and
calculation of Value at Risk.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 60

Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

• Volatility can be measured by the conditional variance of asset
return 𝑌𝑡 given the information available at time 𝑡 − 1:
2
𝜎𝑡2 ≡ var 𝑌𝑡 ∣ 𝐼𝑡−1 = 𝐸 𝑌𝑡 − 𝐸 𝑌𝑡 ∣ 𝐼𝑡−1 ∣ 𝐼𝑡−1
• An example of the conditional variance is the AutoRegressive
Conditional Heteroskedasticity (ARCH) model, originally
proposed by Engle (1982).
Engle's (1982) ARCH(q) model
𝑌𝑡 = 𝜇𝑡 + 𝜀𝑡 ,
𝜀𝑡 = 𝜎𝑡 𝑧𝑡 ,
𝜇𝑡 = 𝐸 𝑌𝑡 ∣ 𝐼𝑡−1 ,
𝑞 2
𝜎𝑡2 = 𝛼 + ∑𝑗=1 𝛽𝑗 𝜀𝑡−𝑗 , 𝛼 > 0, 𝛽 > 0
𝑧𝑡 ∼ 𝐼𝐼𝐷(0,1).
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 61
Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

⚫ 𝐴𝑅𝐶𝐻(𝑞) can explain a well-known stylized fact in financial

markets -- volatility clustering. It can also explain the non-
Gaussian heavy tail of asset returns.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 62

Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

◆ Question
How to estimate a volatility model?

• Here, the models for the conditional mean 𝜇𝑡 and the

conditional variance 𝜎𝑡2 are assumed to be correctly specified,
but the conditional distribution of 𝑌𝑡 is unknown, because the
distribution of the standardized innovation 𝑧𝑡 is unknown.
Thus, the popular maximum likelihood estimation (MLE)
method cannot be used.
• Nevertheless, one can assume that 𝑧𝑡 is 𝐼𝐼𝐷 𝑁(0,1) or
follows other plausible distribution. Under this assumption, we
can obtain a conditional distribution of 𝑌𝑡 given 𝐼𝑡−1 and
estimate model parameters using the MLE procedure.
ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 63
Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

◆ Question
How to estimate a volatility model?

• Although 𝑧𝑡 is not necessarily IID𝑁(0,1) and we know this, the

estimator obtained this way is still consistent for the true model
parameters.
• However, because the conditional distribution of 𝑧𝑡 given 𝐼𝑡−1
is likely to be misspecified, the asymptotic variance of this
estimator is generally larger than that of the MLE (which is based
on the true distribution of 𝑧𝑡 ).
• This is the cost one has to pay not knowing the true distribution
of 𝑧𝑡 .

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 64

Illustrative Examples

Example 1.6 (Cont.): Volatility Clustering and ARCH Models

◆ Question (Cont.)
How to estimate a volatility model?
This method is called the quasi-MLE, or QMLE (see, e.g., White
1994). Inference procedures based on the QMLE are different from
those based on the MLE.
• The popular likelihood ratio test cannot be used. The difference
comes from the fact that the asymptotic variance of the QMLE is
different from that of the MLE, just like the fact that the asymptotic
variance of the OLS estimator under conditional heteroskedasticity
is different from that of the OLS under conditional homoskedasticity.
• Incorrect calculation of the asymptotic variance estimator for the
QMLE will lead to misleading inference and conclusion.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 65

Illustrative Examples

Example 1.7: Modeling Economic Duration

• Suppose we are interested in

✓ the time it takes for an unemployed person to find a job,
✓ the time that elapses between two trades or two price changes,
✓ the time length of a strike,
✓ the time length before a cancer patient dies,
✓ the time length before a financial crisis (e.g., credit default
risk) comes out,
✓ the time length before a startup technology firm goes bankrupt,
✓ the time length before a family gets out of poverty.

Such analysis is called duration analysis or survival analysis.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 66

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

• In practice, the main interest often lies in the question of how long
a duration will continue, given that it has not finished yet. The
so-called hazard rate measures the chance that the duration will
end now, given that it has not ended before. This hazard rate
therefore can be interpreted as the chance to find a job, to trade, to
end a strike, etc.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 67

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

The survival function:

𝑆(𝑡) = 𝑃 𝑇𝑖 > 𝑡 = 1 − 𝐹(𝑡)
and the hazard rate
𝑃 𝑡 < 𝑇𝑖 ≤ 𝑡 + 𝛿 ∣ 𝑇𝑡 > 𝑡 𝑓(𝑡)
𝜆(𝑡) = lim𝛿→0+ =
𝛿 𝑆(𝑡)
where
✓ 𝑇𝑖 is the duration from a population with the probability density
function 𝑓(𝑡) ,
✓ 𝐹(𝑡) is probability distribution function.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 68

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

• Intuitively, the hazard rate 𝜆 𝑡 is the instantaneous probability

that an event of interest will end at time 𝑡 given that it has lasted
for period 𝑡.
• Note that the specification of 𝜆 𝑡 is equivalent to a specification
of the probability density 𝑓 𝑡 .
• But 𝜆 𝑡 is more interpretable from an economic point of view.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 69

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

The hazard rate may not be the same for all individuals. To
control heterogeneity across individuals, we assume that the
individual-specific hazard rate depends on some individual
characteristics 𝑋𝑖 via proportional hazard model (Cox (1972)):
𝜆𝑖 (𝑡) = exp 𝑋𝑖′ 𝛽 𝜆(𝑡)
The parameter
𝜕 1 𝜕
𝛽= ln 𝜆𝑖 (𝑡) = 𝜆 (𝑡).
𝜕𝑋𝑖 𝜆𝑖 (𝑡) 𝜕𝑋𝑖 𝑖

can be interpreted as the marginal relative effect of 𝑋𝑖 on the

hazard rate of individual 𝑖.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 70

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

Inference of 𝛽 will allow one to examine how individual

characteristics affect the duration of interest.

• For example, suppose 𝑇𝑖 is the unemployment duration for

individual 𝑖 then the inference of 𝛽 will allow us to examine how
individual characteristics can affect the unemployment duration.
✓ age,
✓ education,
✓ gender,
✓ ...,
• This will provide important policy implication on labor markets.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 71

Illustrative Examples

Example 1.7 (Cont.): Modeling Economic Duration

The conditional probability density function of 𝑌𝑖 given 𝑋𝑖

𝑓𝑖 (𝑡) = 𝜆𝑖 (𝑡)𝑆𝑖 (𝑡)

where
𝑡
✓ 𝑆𝑖 (𝑡) = exp −∫0 𝜆𝑖 (𝑠)𝑑𝑠 is the survival function.

✓ We can estimate 𝛽 by the maximum likelihood estimation

method.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 72

 CONTENTS

1.1 Introduction
1.2 Quantitative Features of Modern Economics
1.3 Mathematical Modeling
1.4 Empirical Validation
1.5 Illustrative Examples
1.6 Limitations of Econometric Analysis

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 73

Limitations of Econometric Analysis

• Economic data are not produced by a large number of repeated

random experiments, due to the fact that an economy is not a
controlled experiment. Most economic data are nonexperimental
in their nature.

• First, as a simplification of reality, economic theory or model

can only capture the main or most important factors, but the
observed data is the joint outcome of many factors together,
and some of them are unknown and unaccounted for.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 74

Limitations of Econometric Analysis

Some limitations on econometric analysis:

First, as a simplification of reality, economic theory or model

can only capture the main or most important factors, but the
observed data is the joint outcome of many factors together, and
some of them are unknown and unaccounted for.

These unknown factors are well present but

their influences are ignored in economic
modeling. This is unlike natural science,
where one can remove secondary factors via
controlled experiments.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 75

Limitations of Econometric Analysis

Some limitations on econometric analysis:

Second, an economy is an irreversible or non repeatable system. A

consequence of this is that data observed are a single realization
of economic variables.

• To conduct statistical analysis of economic data, economists and

econometricians often assume some time-invarying "common
features" of an economic system so as to use time series data or
cross-sectional data of different economic variables.
• These common features are usually termed as "stationarity" or
"homogeneity" of the economic system.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 76

Limitations of Econometric Analysis

Some limitations on econometric analysis:

Third, economic relationships are often changing over time for an

economy.

Regime shifts and structural changes are rather a rule than an

exception, due to technology shocks and changes in preferences,
population structure and institution arrangements. A well-known
example is the so-called Great Moderation of the American
macroeconomy, whose volatilities have been declining since the
1980s (e.g., Bernakee 2004).

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 77

Limitations of Econometric Analysis

Great Moderation of the American macroeconomy

U.S. GDP growth rate (%)

U.S. inflation rate (%)

Year Year

Figures are the time series plots of annual U.S. GDP growth rates
and U.S. inflation rates, from which one can see that volatilities of
U.S. GDP growth rates and inflation rates have been declining since
mid-1980s.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 78

Limitations of Econometric Analysis

Great Moderation of the Chinese macroeconomy

Chinese GDP growth rate (%)

Chinese inflation rate (%)

Year Year

Similar and perhaps more striking phenomena can also been observed
for Chinese GDP growth rates and inflation rates. Figures show that
the volatilities of Chinese GDP growth rates and inflation rates have
been declining since 1990s.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 79

Roles of Probability and Statistics

⚫ An unstable economic relationship makes it difficult for out-of-

sample forecasts and policy-making.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 80

Limitations of Econometric Analysis

Some limitations on econometric analysis:

Fourth, data quality.

⚫ The success of any econometric study hinges on the quantity as

well as the quality of data.
⚫ However, economic data may be subject to various defects.
✓ The data may be badly measured or may correspond only
vaguely to the economic variables defined in the model.
✓ Some of the economic variables may be inherently
unmeasurable.
✓ Some relevant variables may be missing from the model.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 81

Limitations of Econometric Analysis

Some limitations on econometric analysis:

Moreover, sample selection bias will also cause a problem.

• In China, there may have been a tendency to over-report or

estimate the GDP growth rates given the existing institutional
promotion mechanism for local government officials.
• All these data problems remain even in
the current era of Big Data, which
consists of both structural and
unstructural data. The latter includes data
in form of texts, graphs, voices, videos,
and etc.

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 82

 Thank You !

ADVANCED ECONOMETRICS Introduction to Econometrics September 22, 2024 83